Bug#503268: Another occurrence on lenny

[Tim Small]

A little info to add.

Again, similar messages, and the program which is getting the error
isn't giving anything away as to it's identity:

date    10 July 2011 06:25
subject Cron <root at xeon1> test -x /usr/sbin/anacron || ( cd / && run-parts
--report /etc/cron.daily )

 /dev/dm-1: stat failed: No such file or directory
 Path /dev/dm-1 no longer valid for device(253,1)
 /dev/block/253:1: stat failed: No such file or directory
 Path /dev/block/253:1 no longer valid for device(253,1)
 /dev/disk/by-id/dm-name-vmgroup1-apollo-real: stat failed: No such file or
directory
 Path /dev/disk/by-id/dm-name-vmgroup1-apollo-real no longer valid for
device(253,1)


but this is perhaps relevant - on the same machine I'm seeing this:

xeon1:~# ls -al /dev/disk/by-id/ | grep -i real
lrwxrwxrwx 1 root root   10 2011-07-14 07:34
dm-name-vmgroup1-teucer-real -> ../../dm-1
lrwxrwxrwx 1 root root   10 2011-07-14 07:34
dm-uuid-LVM-8kkkx4Qo1pOv7IsAeAjZP0fNPOqoUPz583J1vuIXFh0WMY4oATuHoxhhf0RYBnf0-real
-> ../../dm-1
xeon1:~# ls -al /dev/disk/by-id/ | grep -i teuc
lrwxrwxrwx 1 root root   11 2011-04-21 12:12 dm-name-vmgroup1-teucer ->
../../dm-25
lrwxrwxrwx 1 root root   10 2011-07-14 07:34
dm-name-vmgroup1-teucer-real -> ../../dm-1
xeon1:~# lvs
  LV       VG       Attr   LSize   Origin Snap%  Move Log Copy%  Convert
  apollo   vmgroup1 -wi-ao  20.00G                                     
  atlas    vmgroup1 -wi-ao  50.00G                                     
  eris     vmgroup1 -wi-ao 100.00G                                     
  gaia     vmgroup1 -wi-ao   5.00G                                     
  gamelia  vmgroup1 -wi-ao   5.00G                                     
  ganymede vmgroup1 -wi-ao  10.00G                                     
  georgos  vmgroup1 -wi-ao   5.00G                                     
  geras    vmgroup1 -wi-ao   5.00G                                     
  geryon   vmgroup1 -wi-ao   5.00G                                     
  glaucus  vmgroup1 -wi-ao   5.00G                                     
  gyges    vmgroup1 -wi-ao   5.00G                                     
  helios   vmgroup1 -wi-ao  10.00G                                     
  hermes   vmgroup1 -wi-ao  30.00G                                     
  nereus   vmgroup1 -wi-ao   5.00G                                     
  p3d1     vmgroup1 -wi-ao  10.00G                                     
  pallas   vmgroup1 -wi-ao   5.00G                                     
  pegasus  vmgroup1 -wi-ao   5.00G                                     
  phoebe   vmgroup1 -wi-ao   5.00G                                     
  proteus  vmgroup1 -wi-ao  30.00G                                     
  pythia   vmgroup1 -wi-ao   9.00G                                     
  tethys   vmgroup1 -wi-ao   1.50G                                     
  teucer   vmgroup1 -wi-ao   5.00G                                     
  thebe    vmgroup1 -wi-ao  20.00G                                     
  xelos    vmgroup1 -wi-ao  20.00G

xeon1:~# dmsetup table  | grep -i teuc
vmgroup1-teucer-real: 0 10485760 linear 9:3 1012851072
vmgroup1-teucer: 0 10485760 linear 9:3 1012851072

- so my reading of this is that there are two identical mappings set up
for some reason, and the "-real" one gets dm-1.  Which would perhaps tie
in with the theory that this is some sort of race?  Why I have this
"-real" mapping left over in this case from yesterday morning's lvcheck
run, I've no idea...

... I realise that the email date doesn't match up with the timestamp on
the device files, but maybe this is a related event happening again?

Tim.